Search

International audience; The world faces complex challenges for chemical hazard assessment. Microfluidic bioartificial organs enable the spatial and temporal control of cell growth and biochemistry, critical for organ-specific metabolic functions and particularly relevant to testing the metabolic dose-response signatures associated with both pharmaceutical and environmental toxicity. Here we present an approach combining a microfluidic system with H-1 NMR-based metabolomic footprinting as a high-throughput small-molecule screening approach. We characterized the toxicity of several molecules: ammonia (NH3), an environmental pollutant leading to metabolic acidosis and liver and kidney toxicity; dimethylsulfoxide (DMSO), a free radical-scavenging solvent; and N-acetyl-para-aminophenol (APAP, or paracetamol), a hepatotoxic analgesic drug. We report organ-specific NH3 dose-dependent metabolic responses in several microfluidic bioartificial organs (liver, kidney, and cocultures), as well as predictive (99% accuracy for NH3 and 94% for APAP) compound-specific signatures. Our integration of microtechnology, cell culture in microfluidic biochips, and metabolic profiling opens the development of so-called "metabolomics-on-a-chip" assays in pharmaceutical and environmental toxicology.

High-mass resolution multi-stage mass spectrometry (MS(n)) fragmentation was tested for differentiation and identification of metabolites, using a series of 121 polyphenolic molecules. The MS(n) fragmentation approach is based on the systematic breakdown of compounds, forming a so-called spectral tree. A chip-based nanoelectrospray ionization source was used combined with an ion-trap, providing reproducible fragmentation, and accurate mass read-out in an Orbitrap Fourier transform (FT) MS enabling rapid assignment of elemental formulas to the molecular ions and all fragment ions derived thereof. The used protocol resulted in reproducible MS(n) fragmentation trees up to MS(5). Obtained results were stable over a 5 month time period, a concentration change of 100-fold, and small changes in normalized collision energy, which is key to metabolite annotation and helpful in structure and substructure elucidation. Differences in the hydroxylation and methoxylation patterns of polyphenolic core structures were found to be reflected by the differential fragmentation of the entire molecule, while variation in a glycosylation site displayed reproducible differences in the relative intensities of fragments originating from the same aglycone fragment ion. Accurate MS(n)-based spectral tree data are therefore a powerful tool to distinguish metabolites with similar elemental formula, thereby assisting compound identification in complex biological samples such as crude plant extracts.

Two major development areas in HPLC-NMR hyphenation are postcolumn solid-phase extraction (HPLC-SPE-NMR) and capillary separations with NMR detection by means of solenoidal microcoils (CapNMR). These two techniques were combined off-line into HPLC-SPE-CapNMR, which combines the advantage of high loadability of normal-bore HPLC columns with high mass sensitivity of capillary NMR probes with an active volume of 1.5 microL. The technique was used for rapid identification of complex sesquiterpene lactones and esterified phenylpropanoids present in an essentially crude plant extract (toluene fraction of an ethanolic extract of Thapsia garganica fruits). Elution profiles of 10 x 1 mm i.d. SPE cartridges filled with poly(divinylbenzene) resin were found to be only marginally broader than those observed upon direct injection of 6-microL samples into the probe. Thus, the technique focuses analytes emerging in the HPLC elution bands of 0.5-1 mL into volumes of approximately 10 microL, compatible with the CapNMR probe. Using this technique, nine natural products (1-9) present in the plant extract in amounts varying from 0.1 to 20% were identified by means of 1D and 2D NMR spectra, supported by parallel HPLC-ESIMS measurements. Therefore, HPLC-SPE-CapNMR should be regarded as an attractive alternative to other applications of CapNMR for mixture analysis.